Marketing forecasting basics: how to predict results before you spend

A forecast is an estimate of future results based on past data and planned changes. In marketing, it usually answers three questions.

1. If we spend this much, how many people will we reach?
2. How many of them will become leads?
3. How many of those leads will become customers, and what will they be worth?

The process is the same at any size. Gather what happened before, decide what will change, do the math and then compare the forecast with what actually happens each month.

The simplest marketing forecasting model multiplies the steps of your funnel together.

Visitors × conversion rate × close rate × average sale = forecast revenue.

Here is a made-up example to show the math. Say your site gets 2,000 visits a month, 2% of visitors fill out a form, you close 25% of those leads and an average customer is worth $1,500. That forecasts 40 leads, 10 customers and $15,000 in revenue. Change any one input and you can see what it does to the result.

Pull those inputs from your own records rather than industry averages. Your analytics shows visits and form fills, and your CRM or sales notes show how many leads became customers. If you only have a few months of history, use what you have and widen your range.

The power of this model is that it shows you where to work. If doubling traffic and doubling your close rate both double revenue, the cheaper of the two is usually the smarter move.

Most businesses have busy and slow months. A forecast that ignores that will look wrong by spring.

The fix is to compare each month with the same month last year rather than the month before. If January is always slow, plan for a slow January and judge your marketing against last January, not December.

Google builds this into its own tools. Its help pages say Keyword Planner's forecasts are refreshed daily, based on the last 7 to 10 days and adjusted for seasonality.

You do not have to build every forecast by hand.

Keyword Planner. Inside Google Ads, Keyword Planner can create a keyword plan and show a forecast of how your chosen keywords might perform. It is a good starting point when you are considering search ads for the first time.

Performance Planner. For accounts already running campaigns, Google describes Performance Planner as a tool to create plans for your advertising spend and assess how changes to campaigns might affect key metrics. It is useful for questions like what happens if you raise a budget by 20%.

GA4 predictive metrics. Google Analytics 4 offers predictive metrics, such as purchase probability and predicted revenue. Predicted revenue, for example, estimates revenue from all purchase key events within the next 28 days from users who were active in the last 28 days. These need enough data to work. Google says that in the last 28 days, over a seven-day period, at least 1,000 returning users must have triggered the relevant condition and at least 1,000 must not have. Smaller sites will not qualify, and that is fine. The funnel math above works at any size.

A single number sounds precise and is almost always wrong. A range is more honest and more useful.

Build three versions.

• Conservative. Your inputs at the low end of what you have seen.
• Expected. Your typical results.
• Optimistic. Your best realistic month.

Then plan your spending and staffing around the conservative case, and treat anything better as a bonus.

Every forecast rests on assumptions, such as a conversion rate holding steady or a new campaign performing like an old one. Others are easy to forget, like a price change, a new competitor nearby or a slower sales team during vacation season. Write them next to the numbers. When the forecast misses, the assumption list tells you why, and that is how the next forecast gets better.

The U.S. Small Business Administration's business plan guide makes the same connection. It tells owners to describe their marketing and sales strategies thoroughly because they will refer to that section later when they make financial projections.

Compare forecast with actual results once a month. Look at each step of the funnel separately. If traffic hit the forecast but leads did not, the problem is on the website or the offer, not the ads.

For more on using your own numbers to steer marketing, read our guides to data-driven marketing and building a marketing strategy from scratch.